How CAMERA's monitor works

We believe transparency is our best defense against the charge that coordinated advocacy is "astroturf." Here is exactly how an alert reaches you.

1. AI first pass. Software reads public coverage and flags stories where a contested claim may be stated as settled fact. It decides only what a human looks at.
2. Two named human gates. A staff analyst verifies the error and approves it under their own name; a second, different staff member approves the exact wording of any volunteer alert before it is published here.
3. One canonical sample. There is one model letter and one post per platform — a recognizable, honest campaign, never machine-generated variations designed to look like independent voices.
4. Institutional contacts only. Letters go to an outlet's published corrections desk — never a journalist's personal details.
5. You act as yourself. The system cannot send, post, or publish anything. You copy, edit into your own words, and send from your own accounts.
6. Published in machine-readable form. Verified corrections and active alerts also carry standard schema.org ClaimReview data, so fact-check aggregators and search engines can read them. That data says only what this page says: CAMERA is the author of the review, the outlet is named as the publisher of the article, and no individual journalist is ever named.

To dispute an alert, contact CAMERA staff. ← Back to alerts